mirror of
https://github.com/tinyhumansai/openhuman.git
synced 2026-07-28 13:32:23 +00:00
Steven EnamakelandGitHub
7bcf43314a
* Add JSON-RPC schema definition and HTTP schema endpoint - Introduced a new `schema.json` file containing detailed definitions for various JSON-RPC methods, including their inputs, outputs, and descriptions. - Implemented a new HTTP endpoint `/schema` in the core server to serve the JSON-RPC schema, enhancing API documentation and accessibility. - Updated the core HTTP router to include the new schema route, improving the overall structure and usability of the API. - Enhanced error handling and response formatting in the server to ensure consistent feedback for schema requests. * Update TypeScript configuration and refactor core RPC client - Changed TypeScript target from ES2020 to ESNext and updated library references in `tsconfig.json` for improved compatibility with modern features. - Refactored `coreRpcClient.ts` to enhance JSON-RPC request handling, including the introduction of legacy method aliases and improved error handling. - Updated API service methods in `authApi.ts` and `channelConnectionsApi.ts` to utilize the new core RPC client structure, streamlining authentication and channel connection processes. - Added new utility functions for managing JSON-RPC requests and responses, improving code organization and maintainability. - Enhanced test coverage for the new RPC client methods and refactored existing tests to align with the updated structure. * Enhance Tauri configuration and refactor daemon program arguments - Updated `tauri.conf.json` to include additional macOS infoPlist settings for better application identification and icon management. - Refactored the `daemon_program_args` function in `common.rs` to improve clarity by renaming the parameter to `_exe`, indicating it is unused. This change enhances code readability and maintainability. * Refactor Tauri configuration by removing unused macOS infoPlist settings - Updated `tauri.conf.json` to streamline macOS configuration by removing unnecessary infoPlist entries while retaining essential settings for the application. - This change enhances clarity and maintainability of the Tauri configuration file. * Enhance authentication flow and testing documentation - Introduced a new `isAuthBootstrapComplete` state in the authentication slice to manage the completion of the authentication bootstrap process. - Updated the `UserProvider` to set the `isAuthBootstrapComplete` state based on the authentication status, improving session restoration logic. - Modified route components (`DefaultRedirect`, `ProtectedRoute`, `PublicRoute`) to conditionally render based on the `isAuthBootstrapComplete` state, enhancing user experience during the authentication process. - Added a comprehensive testing guide in `CLAUDE.md`, detailing unit and E2E testing practices, including setup, authoring rules, and a checklist for test coverage. - Updated the `SettingsHome` component to redirect to the home page instead of the login page upon logout, streamlining user navigation. - Enhanced the `LocalModelPanel` to track download progress and manage local AI assets more effectively, improving overall functionality. * Add resolutions for @tauri-apps/api dependency in package.json - Introduced a resolutions field in package.json to enforce the use of @tauri-apps/api version 2.10.1, ensuring compatibility across workspaces. - Updated dependencies in app/package.json to include @tauri-apps/api version 2.10.1, aligning with the new resolution. - Adjusted yarn.lock to reflect the updated version of @tauri-apps/api, enhancing dependency management and consistency. * Refactor Tauri configuration and enhance E2E build process - Updated `tauri.conf.json` to remove unused resource paths, streamlining the configuration for better maintainability. - Modified `wdio.conf.ts` to improve application path resolution for macOS, allowing for multiple bundle base checks to enhance compatibility. - Refactored `e2e-build.sh` to disable updater artifacts for E2E builds and introduced a conditional cargo clean mechanism, improving build efficiency and clarity. * Enhance E2E testing setup and documentation - Updated `CLAUDE.md` to clarify the default behavior of `OPENHUMAN_WORKSPACE` in `e2e-run-spec.sh`, emphasizing automatic creation and cleanup for reproducible E2E runs. - Modified `e2e-run-spec.sh` to implement automatic temporary workspace creation when `OPENHUMAN_WORKSPACE` is not set, improving usability for debugging and testing. - Enhanced cleanup logic in `e2e-run-spec.sh` to ensure proper removal of temporary workspaces after tests, contributing to a cleaner testing environment. * Implement shared mock backend for testing and enhance documentation - Introduced a shared mock backend for unit and integration tests, allowing for deterministic API behavior across app and Rust tests. - Updated `CLAUDE.md` to include detailed instructions on using the shared mock backend, including key admin endpoints and manual run commands. - Modified `package.json` to add scripts for running the mock API server and Rust tests with the mock backend. - Refactored test setup to utilize the new mock backend, improving test reliability and isolation. - Removed obsolete MSW handlers and server setup, streamlining the testing framework. * Enhance authentication state management and testing coverage - Introduced `isAuthBootstrapComplete` state in the authentication slice to track the completion of the authentication process. - Updated `ProtectedRoute` and `PublicRoute` components to utilize the new state, improving user experience during authentication. - Enhanced test cases for `ProtectedRoute` and `PublicRoute` to reflect the updated authentication state structure. - Added a new end-to-end test for the authentication flow, ensuring proper handling of OAuth tokens and session management. - Improved mock setup in tests to better simulate authentication scenarios, enhancing test reliability and coverage. * Enhance README.md with architecture overview and component roles - Added detailed descriptions of the OpenHuman architecture, highlighting the separation of business logic and UI components. - Explained the roles of Rust and the UI in the monorepo, including the use of JSON-RPC, QuickJS, Vite, React, and Tauri. - Documented the structure of controllers and the RPC surface, emphasizing the shared contract for automation and testing. - Provided links to further documentation for architecture, frontend structure, and Tauri commands. * Integrate ServiceBlockingGate component and enhance loading states - Added the `ServiceBlockingGate` component to manage service availability and display appropriate loading screens. - Updated `DefaultRedirect`, `ProtectedRoute`, and `PublicRoute` components to show a loading indicator while authentication bootstrap is in progress. - Implemented timeout handling in `UserProvider` for improved authentication state management. - Introduced tests for `ServiceBlockingGate` to ensure proper rendering and functionality under various service states. * Implement core RPC URL resolution and default hash route handling - Added a function to resolve the core RPC URL based on the environment, improving flexibility for Tauri and non-Tauri contexts. - Introduced a default hash route handler in the main application entry point to ensure proper navigation behavior. - Updated the core RPC command to expose the resolved RPC URL, enhancing the integration with the frontend. - Refactored the core RPC client to utilize the new URL resolution logic, ensuring consistent API calls. * Refactor backend URL usage to API_BASE_URL - Replaced all instances of BACKEND_URL with API_BASE_URL across various components and services to standardize API endpoint references. - Updated OAuth provider configurations, settings panels, hooks, and services to ensure consistent API calls. - Enhanced test setups to reflect the new API_BASE_URL, improving test reliability and alignment with the updated configuration. * Implement RouteLoadingScreen for improved loading states - Introduced a new `RouteLoadingScreen` component to provide a consistent loading experience across various routes. - Updated `DefaultRedirect`, `ProtectedRoute`, and `PublicRoute` components to utilize `RouteLoadingScreen` while waiting for authentication bootstrap completion. - Enhanced `MiniSidebar` to hide on additional public/setup routes, improving user navigation experience. - Refactored `UserProvider` to streamline authentication state management by removing unnecessary references. * Add feature design workflow section to CLAUDE.md - Introduced a comprehensive workflow for feature design, outlining steps from specification to UI implementation and testing. - Emphasized the importance of grounding designs in existing codebases and defined planning rules for E2E scenarios. - Provided detailed instructions for implementing features in Rust, conducting JSON-RPC tests, and building UI components in the Tauri app. * Refactor backend URL handling and improve OAuth flow - Replaced static API_BASE_URL references with dynamic backend URL resolution across various components and services, enhancing flexibility for Tauri and non-Tauri environments. - Updated OAuth provider configurations to utilize the new backend URL logic, ensuring consistent login URL generation. - Refactored API client and socket service to fetch the backend URL dynamically, improving reliability in different deployment contexts. - Introduced a new service for resolving the backend URL, streamlining the configuration and enhancing test setups. * Add debug logging guidelines to CLAUDE.md - Introduced comprehensive guidelines for implementing development-oriented debug logging in both Rust and the app. - Emphasized the importance of logging at appropriate levels (`debug`/`trace`) and following existing patterns for consistency. - Provided instructions on avoiding sensitive information in logs and ensuring terminal output is grep-friendly for easier debugging during development. * Enhance service management with new mock functionality and E2E tests - Added a mock service manager to facilitate deterministic service behavior during end-to-end tests, enabling better simulation of service states. - Introduced new buttons in the `ServiceBlockingGate` component for restarting and uninstalling services, improving user control over service management. - Implemented a comprehensive E2E test suite for the service connectivity flow, covering installation, starting, stopping, restarting, and uninstalling services. - Updated package scripts to include a new E2E test for service connectivity, enhancing testing coverage and reliability. - Refactored service operations to support mock functionality, ensuring consistent behavior across testing and production environments. * Enhance ServiceBlockingGate with improved logging and periodic health polling - Introduced periodic health polling in the `ServiceBlockingGate` component to refresh service status every 3 seconds, enhancing responsiveness to service state changes. - Added detailed logging for various operations, including service status checks and error handling, to improve traceability and debugging. - Updated E2E tests to include logging steps for better visibility during service connectivity flow tests. - Refactored error handling to ensure consistent logging of error messages across service operations. * Refactor E2E testing scripts and enhance CLAUDE.md documentation - Updated paths in CLAUDE.md to reflect the new location of the E2E run script, ensuring accurate instructions for running tests. - Removed outdated E2E test scripts from package.json and migrated relevant functionality to app/package.json for better organization. - Introduced new E2E testing scripts for specific flows (auth, login, payment, etc.) to streamline testing processes and improve modularity. - Added a script to run all E2E flows sequentially, enhancing test coverage and simplifying execution. - Improved documentation for E2E testing procedures in CLAUDE.md, providing clearer guidance for developers. * Refactor imports and enhance code readability - Removed duplicate import of `ServiceBlockingGate` in `App.tsx` for cleaner code. - Improved readability in `MiniSidebar.tsx` by formatting conditional statements. - Reordered imports in `PublicRoute.tsx` for consistency. - Enhanced formatting in `ServiceBlockingGate.tsx` for better clarity in asynchronous operations. - Streamlined import statements in various test files and components for improved organization. - Updated `LocalModelPanel.tsx` to enhance button disable logic readability. - Refactored CORS headers in `jsonrpc.rs` for better formatting. * ci: align build and test workflows with app workspace e2e * ci: align release and typecheck workflows with current workspace * Enhance ServiceBlockingGate functionality with improved refresh options - Introduced a new `RefreshOptions` type to customize the behavior of the `refreshStatus` function, allowing for conditional error clearing and status checking. - Updated the `refreshStatus` function to utilize the new options, enhancing control over service state updates. - Modified the button click handler to pass the new options, improving user experience during service refresh operations. - Adjusted state updates to prevent unnecessary re-renders and maintain consistency in service state management. * Refactor RefreshOptions type for improved clarity in ServiceBlockingGate - Consolidated the definition of the `RefreshOptions` type into a single line for better readability. - Maintained existing functionality while enhancing code clarity in the `ServiceBlockingGate` component.
94 lines
3.0 KiB
Bash
Executable File
94 lines
3.0 KiB
Bash
Executable File
#!/usr/bin/env bash
|
|
#
|
|
# Run a single WebDriverIO E2E spec (Appium mac2 + mock server in spec).
|
|
#
|
|
# Usage:
|
|
# ./app/scripts/e2e-run-spec.sh test/e2e/specs/login-flow.spec.ts [log-suffix]
|
|
#
|
|
set -euo pipefail
|
|
|
|
SPEC="${1:?spec path required}"
|
|
LOG_SUFFIX="${2:-$(basename "$SPEC" .spec.ts)}"
|
|
|
|
APPIUM_PORT="${APPIUM_PORT:-4723}"
|
|
E2E_MOCK_PORT="${E2E_MOCK_PORT:-18473}"
|
|
|
|
SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
|
|
APP_DIR="$(cd "$SCRIPT_DIR/.." && pwd)"
|
|
REPO_ROOT="$(cd "$APP_DIR/.." && pwd)"
|
|
cd "$APP_DIR"
|
|
# shellcheck source=/dev/null
|
|
source "$SCRIPT_DIR/e2e-resolve-node-appium.sh"
|
|
|
|
CREATED_TEMP_WORKSPACE=""
|
|
APPIUM_PID=""
|
|
if [ -z "${OPENHUMAN_WORKSPACE:-}" ]; then
|
|
OPENHUMAN_WORKSPACE="$(mktemp -d)"
|
|
CREATED_TEMP_WORKSPACE="$OPENHUMAN_WORKSPACE"
|
|
export OPENHUMAN_WORKSPACE
|
|
echo "Using temporary OPENHUMAN_WORKSPACE: $OPENHUMAN_WORKSPACE"
|
|
else
|
|
echo "Using OPENHUMAN_WORKSPACE from environment: $OPENHUMAN_WORKSPACE"
|
|
fi
|
|
|
|
if [ "${OPENHUMAN_SERVICE_MOCK:-0}" = "1" ] && [ -z "${OPENHUMAN_SERVICE_MOCK_STATE_FILE:-}" ]; then
|
|
OPENHUMAN_SERVICE_MOCK_STATE_FILE="$OPENHUMAN_WORKSPACE/service-mock-state.json"
|
|
export OPENHUMAN_SERVICE_MOCK_STATE_FILE
|
|
echo "Using OPENHUMAN_SERVICE_MOCK_STATE_FILE: $OPENHUMAN_SERVICE_MOCK_STATE_FILE"
|
|
fi
|
|
|
|
cleanup() {
|
|
if [ -n "$APPIUM_PID" ]; then
|
|
echo "Stopping Appium (pid $APPIUM_PID)..."
|
|
kill "$APPIUM_PID" 2>/dev/null || true
|
|
wait "$APPIUM_PID" 2>/dev/null || true
|
|
fi
|
|
if [ -n "$CREATED_TEMP_WORKSPACE" ]; then
|
|
rm -rf "$CREATED_TEMP_WORKSPACE"
|
|
fi
|
|
}
|
|
trap cleanup EXIT
|
|
|
|
export VITE_BACKEND_URL="http://127.0.0.1:${E2E_MOCK_PORT}"
|
|
export BACKEND_URL="http://127.0.0.1:${E2E_MOCK_PORT}"
|
|
|
|
echo "Cleaning cached app data..."
|
|
rm -rf ~/Library/WebKit/com.openhuman.app
|
|
rm -rf ~/Library/Caches/com.openhuman.app
|
|
rm -rf "$HOME/Library/Application Support/com.openhuman.app"
|
|
|
|
DIST_JS="$(ls dist/assets/index-*.js 2>/dev/null | head -1)"
|
|
if [ -z "$DIST_JS" ]; then
|
|
echo "ERROR: No frontend bundle found at dist/assets/index-*.js." >&2
|
|
echo " Run 'yarn test:e2e:build' to build the app before running E2E tests." >&2
|
|
exit 1
|
|
fi
|
|
if ! grep -q "127.0.0.1:${E2E_MOCK_PORT}" "$DIST_JS"; then
|
|
echo "ERROR: frontend bundle does NOT contain mock server URL (127.0.0.1:${E2E_MOCK_PORT})." >&2
|
|
echo " Run 'yarn test:e2e:build' to rebuild with the mock URL." >&2
|
|
exit 1
|
|
fi
|
|
echo "Verified: frontend bundle contains mock server URL."
|
|
|
|
APPIUM_LOG="/tmp/appium-e2e-${LOG_SUFFIX}.log"
|
|
NODE_VER=$("$NODE24" --version)
|
|
echo "Starting Appium on port $APPIUM_PORT (Node $NODE_VER)..."
|
|
echo " Appium logs: $APPIUM_LOG"
|
|
"$APPIUM_BIN" --port "$APPIUM_PORT" --relaxed-security > "$APPIUM_LOG" 2>&1 &
|
|
APPIUM_PID=$!
|
|
|
|
for i in $(seq 1 30); do
|
|
if curl -sf "http://127.0.0.1:$APPIUM_PORT/status" >/dev/null 2>&1; then
|
|
echo "Appium is ready."
|
|
break
|
|
fi
|
|
if [ "$i" -eq 30 ]; then
|
|
echo "ERROR: Appium did not start within 30 seconds." >&2
|
|
exit 1
|
|
fi
|
|
sleep 1
|
|
done
|
|
|
|
echo "Running E2E spec ($SPEC)..."
|
|
npx wdio run test/wdio.conf.ts --spec "$SPEC"
|